Ink jet printing has become recognized as a prominent contender in the digitally controlled, electronic printing arena because, e.g., of its non-impact, low-noise characteristics, its use of plain paper, and its avoidance of toner transfer and fixing. Ink jet printing mechanisms can be categorized by technology as either drop on demand ink jet (DOD) or continuous ink jet (CIJ).
The first technology, “drop-on-demand” (DOD) ink jet printing, provides ink drops that impact upon a recording surface using a pressurization actuator, for example, a thermal, piezoelectric, or electrostatic actuator. One commonly practiced drop-on-demand technology uses thermal actuation to eject ink drops from a nozzle. A heater, located at or near the nozzle, heats the ink sufficiently to boil, forming a vapor bubble that creates enough internal pressure to eject an ink drop. This form of inkjet is commonly termed “thermal ink jet (TIJ).”
The second technology commonly referred to as “continuous” ink jet (CIJ) printing, uses a pressurized ink source to produce a continuous liquid jet stream of ink by forcing ink, under pressure, through a nozzle. The stream of ink is perturbed using a drop forming mechanism such that the liquid jet breaks up into drops of ink in a predictable manner. One continuous ink jet printing technology uses thermal stimulation of the liquid jet with a heater to form drops that eventually become print drops and non-print drops. Printing occurs by selectively deflecting one of the print drops and the non-print drops and catching the non-print drops. Various approaches for selectively deflecting drops have been developed including electrostatic deflection, air deflection, thermal deflection, mechanical deflection, deflection by alteration of the fluid velocity field, both within the body of ink and externally coupled to the body of ink, and deflection based on changes in the contact free energy of the ink contacting a solid surface (often referred to as surface deflection, as is known in the art of continuous inkjet printing.
In both drop on demand and continuous ink jet technologies print drops land at various positions on the receiver, the potential landing locations of the printed drops can be described by a hypothetical ‘pixel grid’ on the receiver. The representation of the potential landing locations of printed drops as a hypothetical pixel grid is used extensively in technical analyses and in product specifications including printer resolution. For example, as is well known in the art of ink jet printing, in binary printing, each pixel grid on the receiver receives either one or no ink drop. Also by way of example, many products are known in the art of commercial printing having pixel grids of from 600×600 pixels per inch to 2400×2400 pixels per inch. The concept of a pixel grid allows classification of system architectures and is particularly useful in analyzing the effects of drop steering on printer performance. The hypothetical pixel grid on the receiver (or paper or substrate) has a spatial density, typically measured in units of inverse inches (per inch), in both a direction perpendicular and parallel to the direction of the receiver (paper) path relative to the printhead mechanism. These spatial densities are equivalent to the reciprocals of the spatial dimensions in the two directions. Typically, the edge locations for the spatial grid in the direction perpendicular to the receiver path are the same, whereas the edge locations of the spatial grid in the direction of the receiver path can vary up to the spatial dimension in that direction. For example, when even and odd printhead nozzles fire exactly out of phase, the edge locations of the spatial grid in the direction of the receiver path alternate by half the dimension of the spatial grid in that direction, as can be appreciated by one skilled in inkjet systems engineering.
The dimensions of the pixel grid and the way in which drops can fill the pixels of the pixel grid depend on the type of printing system architecture, which in turn is based on the type of drop ejection technology. Continuous drop ejection technologies typically include one or more jetting modules having a plurality of nozzle plates each with nozzles formed in a regular linear array and oriented approximately perpendicular to a receiver path. The nozzles have a well-defined spacing along the nozzle array (typically perpendicular to the receiver path) and hence have a well defined ‘native’ spatial nozzle density measured as the number of nozzles per inch (npi) along the direction of the array. For example, products are known in the art of commercial printing having ‘native’ spatial nozzle densities in the range of 200 to 2400 npi. Typically, each nozzle can print drops onto the receiver.
The pixel grid characterizing the location of the drops printed on the receiver is typically a regular array (that is, evenly spaced in both directions) characterized by a well defined number of pixels per inch perpendicular to the receiver path (usually called the slow scan direction or the direction aligned along the nozzle array) and a well defined number of pixels per inch in the direction of the receiver path (usually called the fast scan direction or the direction aligned perpendicular to the nozzle array). As is well known, the simplest pixel grid is an array of squares with edges aligned (or collinear) in both the direction of the travel path of the receiver and in the direction perpendicular to the travel path of the receiver. However other printing architectures are well known, having, for example, pixel grid arrays of rectangles. This occurs when the receiver speed and drop print frequency are such that the pixel grid in the direction of the receiver travel path is larger than or smaller than (but not equal to) the pixel grid in the direction perpendicular to the receiver path. If the nozzle array is perpendicular to the receiver path and all printed drops fire simultaneously along the nozzle array, than the pixel grid is an array of rectangles with edges aligned in both the direction of the travel path of the receiver and in the direction perpendicular to the travel path of the receiver. If the printed drops fire at delayed times with respect to one another along the nozzle array, than the pixel grid is an array of trapezoids, as is well known in the art of ink jet systems architectures. For binary printing, the vertices of the hypothetical pixel grid array have the same spatial pattern as the landing sites of drops when all drops are printed. Unless stated otherwise, the preferred landing location of drops for binary printing is here taken to be in the center of the pixels of the pixel grid. Printed drops land in the areas defined by the pixel grid (referred to as pixels) in different ways depending on print system architecture. For example, as is well known in the art of ink jet printing, in binary printing, each pixel grid on the receiver receives either one ink drop or no ink drops; where as in contone printing, each pixel receives either a varying number of drops, including zero, or a drop of a variable size.
The spatial density of the pixel grid in the slow and fast scan directions is frequently identical and equal to the native spatial nozzle density. For example pixel grids of 600×600 pixels per inch (often called dots per inch, particularly when referring to binary printers) printers using 600 npi nozzle arrays are know in the art. Here, the first number indicates the spatial density perpendicular to the receiver path and the second indicates the spatial density parallel to the receiver path. However, in some alternate printer system architectures, the spatial density in the fast scan direction is configured to differ very significantly from that in that slow scan direction. For example, printing with a 600 npi nozzle array onto a 600×900 pixels per inch (ppi) grid achieves a different result than from printing with a 600 npi nozzle array onto a 600×600 pixels per inch grid. The 600×900 pixels per inch grid architecture is frequently achieved by moving the receiver 50% slower than in the case of printing on a 600×600 pixels per inch grid, resulting in 50% lower print system productivity but with superior image quality. A 600×900 pixels per inch pixel grid can also be achieved by increasing the frequency of drop formation, but this requires a higher frequency performance of the jetting module and may also require adjusting the drop size so as to avoid excess drop overlap. Such system architectures are useful in product lines that serve different applications, each having different speed and quality requirements.
As another example, in an alternate printer system architecture, the spatial density in the slow scan direction significantly exceeds the native npi of the nozzle array. Prior art teaches the use of a nozzle to address multiple pixels, by steering the drops at least in a direction partially aligned with the nozzle row (perpendicular to the paper path), for the purpose of reducing the number of nozzles required, a nearby nozzle being steered to “cover” drop printing when needed into adjacent pixels. In these cases, nozzles are associated with more than one pixel. Such system architectures can be achieved by steering drops from each nozzle so that each nozzle can print sequentially into multiple, closely adjacent (in the direction perpendicular to the receiver path) pixels. For example, printing with a 200 npi nozzle array onto a pixel grid of 600 pixels per inch in the direction perpendicular to the receiver path can be achieved by having each nozzle print sequentially into three pixels. This results in an increase of image quality due to the higher resolution perpendicular to the receiver path, albeit at a reduction of three in speed, since the receiver must move more slowly to allow time for each nozzle to print in multiple locations.
As another example, in alternate printer system architecture, the spatial density in the slow scan direction is increased in comparison with the native nozzle density by angling the printhead so that the row of nozzles is no longer perpendicular to the receiver path. For example, printing with a 600 npi nozzle array onto a pixel grid of 850 pixels per inch in the direction perpendicular to the receiver path can be achieved by rotating the print module by approximately 45 degrees. Of course, this requires a mechanically precise rotation means, and the resulting module occupies more space in the direction of the receiver path, which adds complexity and cost.
As another example, in an alternate printer system architecture, the spatial density in the fast scan direction is decreased in comparison to that in the slow scan direction. For example, printing with a 600 npi nozzle array onto a pixel grid of 600×300 pixels per inch (300 pixels per inch in the direction along the receiver path) in comparison to a pixel grid of 600×600 pixels per inch can be achieved by doubling the speed of the receiver while keeping the drop formation rate the same, hence increasing productivity.
Typically, most methods of producing inkjet printer systems result in printers having receiver pixel grids fixed at the time of manufacture, for example a pixel grid of size 1200 by 1200 pixels per inch in directions perpendicular and parallel with the receiver path respectively is common, as is a grid size of 600 by 600 pixels per inch. Grid dimensions are often the same, machine to machine. An inkjet printer could be manufactured with an unusual pixel grid density, for example 673 by 1333 pixels per inch, by building nozzles plates with specially spaced nozzles and by running the printer at non-conventional ratios of print frequency to receiver speed. Although, as discussed below, there would be performance advantages to such unusual pixel grid densities, such low volume products are expensive and have not found widespread use.
In some printer system architectures, including binary and contone, the position of drops within receiver pixels can be selectively controlled to improve image quality, for example to improve the accuracy of certain printed characters, such as serifs on individual letters. In this architecture, the position of drops within receiver pixels must be changed very frequently (up to the pixel print rate) since the image content can change from pixel to pixel. Since data flow rates are limited in practice by cost and technology constraints, the number of positions of drops within receiver pixels to improve printed characters is limited.
The receiver pixel dimension perpendicular to the receiver path is generally taught to be constant over the entire length of the printhead for reasons of consistency of image quality and to simplify image data ripping and rasterization. Thus a printing system having a pixel density in the direction perpendicular to the receiver path of 600 pixels per inch along a portion of the printhead generally maintains this density over the entire printhead length. Also, the receiver pixel dimension perpendicular to the receiver path is generally taught to be constant over time during printing. A conventional printer having a particular pixel density in the direction perpendicular to the receiver path is not reconfigurable during printing to a printer having a different pixel density even though there are situations where such pixel density reconfigurations during printing operations would be of value.
Watermarking, for example, is commonly used in secure document printing with one implementation including the encoding of machine readable information in the patterns of printed dots. Typically, watermarking is achieved by subtle variations of the positions of printed drops, although reading this information requires sophisticated image scanners. As such, there remain barriers, including cost and complexity, to reliably printing high quality secure documents and there is a well-recognized need for improvement in this area.
In technologies for watermarking inkjet prints, an important objective is to allow rapid and low cost machine identification or tagging of document origin. Another objective is to prevent copying unauthorized documents inexpensively. For example, it is not difficult to copy documents convincingly using inkjet printing, since both the original print and the copy often have identical or commensurate pixel grids. For example, contone copying machines having high grid densities, for example 1200 (or 2400) pixels per inch, can be operated as machines having grid densities of 600 pixels per inch, simply by omitting print drops in every other (or every fourth) pixel and printing larger drops in the pixels used.
The pixel spacing in the direction parallel to the receiver path is relatively easy to alter, by adjusting the receiver speed. However, this can be done both for the copy machine as well as for the original printer and so does not provide a means of securing documents against copying. Other more complex methods of image water marking have been developed to help prevent unauthorized copying, but such software techniques can be mimicked if the copy printer and original document printer are physically similar. On the other hand, the pixel spacing in the direction perpendicular to the receiver path has not proved easy to alter, although the ability to alter this parameter on the original document printer would present great difficulties for printers attempting to make convincing copies. As noted, an inkjet printer could be manufactured with an unusual pixel grid density, for example 673 by 1333 pixels per inch which would present difficulties of reproduction for machines capable of printing only fixed, standard pixel grids, for example 1200 by 1200 pixels per inch. However, such ‘one-off’ production examples are not cost effective.
A second prior art method to accomplish an altered pixel spacing in the direction perpendicular to the receiver path is available to printers having an array of nozzles each of which can addresses multiple closely adjacent pixels. For example, if each nozzle can address three pixels, then each nozzle could be programmed to address 2 pixels or possibly 4 pixels depending upon the maximum amount of steering available. This type of change in the pixel density in the direction perpendicular to the receiver path very substantially alters the amount of drop steering required and the number to times the drops are steered. The altered pixel density would differ by a large amount from the original density and the result of changing the pixel density would easily be visible to the human eye. In the above example, such alterations would result in a new pixel grid whose spatial density in the direction perpendicular to the receiver was altered by factors of 1.33 and 0.67. These changes would substantially alter the image quality and speed of the printer hence it is not surprising that such alteration is not found in practice. Additionally, an array of nozzles, each of which can address multiple closely adjacent pixels, has a correspondence between nozzles and pixels that is not one to one. This introduces additional cost and system complexity and reduces speed. Alterations resulting in a new pixel grid whose spatial density in the direction perpendicular to the receiver has been increased by a small amount, for example 1%, are not contemplated in the prior art of nozzles which do not have a one to one correspondence between nozzles and receiver pixel.
The representation of the potential landing locations of printed drops as a hypothetical receiver pixel grid is useful in analyzing drop placement errors on the receiver. For example, in binary printing, the printed drops typically are intended to land in the pixel centers, or, if the landing locations are subject to random fluctuations, the mean positions of drops are typically intended to be in the pixel centers. Deviations from the desired position may be measured and corrected in some print system architectures. This is an important image quality issue, since repetitive errors in the position of a single misdirected drop are high visible to the eye. For example, if one nozzle is persistently misdirected and produces drops landing at the bottom right of its intended pixel, image quality is compromised. Corrective steering can be applied to move such drops towards the pixel center and requires only a onetime adjustment. However, in this example, if the nozzle fails entirely, for example, by no longer emitting liquid, then it is generally not possible to correct the operation of that nozzle. This is a common occurrence among drop on demand printers of the thermal inkjet type, and is typically solved by redundancy, i.e. by employing an additional set of nozzles to place drops in the positions the failed nozzle would have placed them, albeit at a different moment in time, or by multiple scans. This procedure is disadvantageous because it slows printer operation in the case the printhead makes many passes over the same receiver area or requires a backup set of nozzles that add cost and complexity. Accordingly, there is a need for an improved solution for failed nozzles, especially for single pass printers in which the document passes only a single time under the printhead.
The concept of potential landing locations of printed drops on a grid can also be extended to analyze drop placement on the catcher for the case the printer is of the continuous type. For example, in binary printing, the non-printed drops typically are intended to land in a particular position on the catcher; when no drops are printed and all drops land on the catcher, the landing positions should ideally form a straight ‘catch line’, with the positions of the drops approximately evenly spaced in the direction along the nozzle array. Deviations from the desired positions are well known to decrease system reliability due to exceptionally non-uniform accumulation of fluid on the catcher, which is particularly severe when the fluid is viscous, as is often the case for inkjet printing inks. Typically, deviations are not controlled; rather printheads are selected to have the best catch performance, for example, those selected for production might have a small root mean square (rms) deviation of the landing locations from the ideal catch line. This approach tends to be costly. As such, there is a need to improve the consistency of landing positions of unprinted drops on a catcher during printing such that these landing positions are as close as possible to desired landing positions during printing.
The representation of the potential landing locations of printed drops as a pixel grid is also useful in compensating for deformations of the receiver, for example deformations due to wet load as subsequent colors are printed. Generally, as is well known in the art of inkjet printing, a high liquid content causes the receiver to stretch, thereby very slightly altering the effective pixel spacing, for example by less than one percent, when an image is printed on a stretched receiver that subsequently dries and returns to its original dimension. If the stretching is uniform, then in the direction along the paper path, the final printed receiver grid can be controlled in principal by altering the receiver speed or the print frequency, so that the dried receiver displays the intended pixel grid in the direction of the receiver path, as is well known. However, this technique cannot be used to keep the intended pixel grid constant in the direction perpendicular to the receiver path because timing cannot alter the pixel grid in that direction and the dried print will exhibit printed drops more closely spaced than desired, as is also well known. Current printers can alter the image data in response to anticipated changes in receiver dimensions, and while this may improve image quality it is not a totally satisfactory solution, since the spacing of drops in the direction perpendicular to the paper path is not restored to the desired values. A need exists, therefore, to guard against image artifacts due to stretching of the receiver.